Strategies for Improving Memory in Students

Students generally have little to no understanding of the memory processes, resulting in an inability to study effectively. To perform better, it is imperative to take control of memory by understanding how memory works, how information can be committed to memory, and how forgetting can be avoided. Some insight into the cognitive load theory can also be applied to improve learning.

Topics in Qualitative Randomized Response Modeling

A growing interest of researchers from diverse fields of inquiry, interlocking social sciences, marketing research and pattern recognition, can be witnessed in study of hidden aspects of individuals. The motivation of this thesis is consistent with the urge of attaining more valid data in survey-based research but with prime focus of quantification of sensitive attributes prevalent in a population. The objectives of facilitating the efforts of exploring stigmatized behaviors through surveys are met by pursuing the well celebrated randomized response approach of gathering information. Emphasizing on highly praised Kuk (1990) model, we propose suits of methods, reinforced with essential features of practical feasibility and capability of ensuring respondents’ privacy. Throughout this thesis, we report our findings in comparison with various contemporary techniques while considering variety of design parameters. A rich account of comparative performance evaluation of the proposed methods, defining the superior performance of the propositions with respect to contemporary models, is also provided. Recognizing the utility of complex sampling designs in practical research, we also offer the extensions of our work for stratified sampling design by suggesting a novel approach combining the responses from direct and indirect questioning approaches. Furthermore, applicability of randomized response approach is extended to generate masked paired comparisons data by proposing a new hybrid domain while taking clues from classic paired comparison model of Bradly and Terry (1952). Moreover, in this research we also explore strategies to pursue the statistical modeling of masked binary responses at multi-level by anticipating randomization procedure behind the extent of masking in the data. The applications of our propositions are delineated by using classic and vibrant data sets.